New pediatric treatment guidelines recommend starting all HIV-infected infants on antiretroviral treatment regardless of their immunologic or clinical profile. Treatment initiation with Lopinavir/ritonavir (LPV/r) is recommended because of concerns about resistance to non-nucleoside reverse transcriptase-inhibitors (NNRTI) following use of nevirapine (NVP) in prevention of mother-to-child HIV transmission (PMTCT). Guidelines offer no specific advice about whether LPV/r-based therapy (usually recommended as a second- line regimen but here recommended as first-line) should be continued life-long for all HIV-infected children starting therapy at a young age. There are several risks associated with indefinite, long-term use of LPV/r- based therapy, including its poor palatability (raising adherence challenges in toddlers and older children), interactions with rifampicin used for co-treatment for tuberculosis, the lack of any suitable second-line regimens, and uncertainty about its long-term metabolic toxicities when used in developing children. We propose an unblinded randomized clinical trial to evaluate a simplification, protease-inhibitor (PI)-sparing treatment strategy among NVP-exposed HIV-infected children treated initially with LPV/r. The primary objective is to test, among children suppressed on LPV/r-based therapy, whether the durability of viral suppression is equivalent when children are switched to efavirenz (EFV)-based therapy. EFV-based therapy is an attractive alternative as it is already recommended for treatment of children >3 years, is widely used, palatable with once daily pediatric formulations, a low, well-described toxicity profile, and is recommended for co-treatment with rifampicin. We propose to recruit 300 HIV-infected children aged 3 to 5 years at a clinical site in Johannesburg, South Africa. Inclusion criteria will include exposure to NVP as part of PMTCT, initiation of LPV/r-based therapy in the first 36 months of life and a viral load < 50 copies/ml. These children will be randomized to either substitute EFV for LPV/r or to continue on their LPV/r-based regimen. Children will be followed with regular viral load and other clinical tests for 48 weeks after randomization. Children in the experimental arm who have breakthrough viremia on the EFV-based regimen will promptly reinitiate the LPV/r regimen. Secondary aims include comparison of immune preservation, toxicities, selection of resistance mutations, and adherence across the two arms. Antiretroviral drug concentrations and adherence will be investigated as possible explanations for the effects of this simplification regimen. The overall goal of the study is to contribute to the evidence base to allow expansion of treatment options for HIV-infected children in low resource settings.